Aware is an immersive generative artwork built around an interactive 3D word cloud displayed in an infinity-mirror enclosure. It transforms text into a spatial visualization and invites viewers to reflect on how artificial intelligence can reshape relationships between knowledge, memory, and perception.
- Extract text from PDF lecture notes.
- Split sentences into individual words.
- remove repeated or low-information words.
- Embed the remaining words into a high-dimensional feature space.
- Reduce the embeddings to three dimensions.
- Render and navigate the resulting word cloud in Processing.
The repository retains separate Python scripts for the preprocessing stages and Processing sketches for the final visualization.
pdf to csv files.py PDF text extraction
divide wording.py Token preparation
filter out useless words.py Word filtering
convert 300d data to 3d.py Dimensionality reduction
embedded_photo.csv Intermediate embedding data
wordcloud/wordcloud.pde Main Processing sketch
wordcloud/no_text/no_text.pde Alternate no-text study
wordcloud/*.csv Prepared 3D datasets
- Install Processing.
- Install the PeasyCam library through Processing's Contribution Manager.
- Clone or download this repository.
- Open
wordcloud/wordcloud.pde. - Keep the referenced CSV files inside the sketch folder.
- Run the sketch.
The preprocessing scripts were developed as an artwork pipeline and may contain machine-specific paths. Review and replace those paths before regenerating the data.
Adjust point scale, camera settings, colours, text size, and connection thresholds in the Processing sketch. Substitute your own corpus by running it through the same extraction, filtering, embedding, and dimensionality-reduction stages.
See the project notes for additional background.
Aware is preserved here as an artwork and research prototype. Reproducing the physical experience also requires the original display and infinity-mirror installation.